Temporal and spatial variation of N2O production from estuarine and marine shallow systems of Cadiz Bay (SW, Spain)

Science of the Total Environment 607–608 (2017) 141–151

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Temporal and spatial variation of N2O production from estuarine and marine shallow systems of Cadiz Bay (SW, Spain) Macarena Burgos ⁎, Teodora Ortega, Jesús M. Forja Dpto. Química-Física, INMAR, Universidad de Cádiz, Campus Río San Pedro, s/n, Puerto Real, Cádiz, Spain

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Dissolved N2O distribution is affected by anthropogenic activities, lateral inputs from salt marshes or river discharges. • The variability of dissolved N2O is associated with the precipitation regime. • Nitrification has been identified as an important process for N2O production within the water column. • The study area acts as a net source of N2O to the atmosphere.

a r t i c l e

i n f o

Article history: Received 27 January 2017 Received in revised form 20 June 2017 Accepted 3 July 2017 Available online xxxx Keywords: N2O Coastal systems Nitrification Air-water flux Spatio-temporal variability

a b s t r a c t There is still much uncertainty regarding the global oceanic emissions of N2O, and particularly emissions from coastal regions, because spatio-temporal datasets have limited coverage. The concentration of dissolved N2O in surface waters and the associated fluxes to the atmosphere have been studied in three coastal systems located near Cadiz Bay (southwestern coast of Spain) over different time scales. The three systems present different hydrodynamic characteristics (an estuary and two marine systems) that influence the distribution of N2O in the water column. Nutrients, oxygen, and particulate organic nitrogen were also measured to investigate the processes responsible for N2O production in the water column. Data on dissolved N2O has been obtained in each system from i) two-year monitoring at fixed station; ii) four seasonal samplings along the longitudinal length of the system; and iii) daily sampling in summer. The concentration of N2O ranges between 1.1 and 292.0 nM indicating very high spatio-temporal variability. In general, the concentration of N2O increased during the rainy season associated with the precipitation regime that, in turn, increases the lateral inputs of organic matter and nutrients from both natural sources (discharges into rivers and adjacent marshes) and anthropogenic activities (agriculture, urban effluents and fish farming). Dissolved N2O also varied with the tides: the highest concentrations were measured during the ebb, which suggests that the systems export N2O to the Bay and adjacent Atlantic Ocean. In addition nitrification seems to be an important process for N2O formation in the water column, which also explains some of the variability in the dataset. The mean atmospheric flux of N2O reveals that entire study area was a net source of N2O to the atmosphere. The fluxes ranged between 0.5 and 313.2 μmol m−2 day−1 in the estuarine system, and between −7.2 and 97.8 μmol m−2 day−1 in the two marine systems. © 2017 Published by Elsevier B.V.

⁎ Corresponding author. E-mail addresses: [email protected] (M. Burgos), [email protected] (T. Ortega), [email protected] (J.M. Forja).

http://dx.doi.org/10.1016/j.scitotenv.2017.07.021 0048-9697/© 2017 Published by Elsevier B.V.

142

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

1. Introduction Nitrous oxide (N2O) is a potent greenhouse gas with a global warming potential (cumulative radiative forcing) 296 times that of CO2 in a 100-year time horizon (IPCC, 2013). Its atmospheric concentration has increased up to 20% above the pre-industrial level. Currently, it contributes about 6% of the overall global warming effect, and of that, the contribution from the agricultural sector is about 16% (Dalal et al., 2003). Therefore, it is important to identify the diverse natural sources of N2O in order to understand their contribution to the overall global N2O budget. Large sinks and sources of N2O in coastal ecosystems have been identified; they include production in sediments (Chong et al., 2012), input from tidal flats (Middelburg et al., 1995), and emission to the atmosphere (Bange, 2006). In marine environments, N2O is mainly produced as a by-product in the first step of autotrophic nitrification (the microbial oxidation of ammonium to nitrate, via hydroxylamine and nitrite), and as an obligate intermediate during anaerobic denitrification (the reduction of microbial nitrate to dissolved gaseous nitrogen) (Bange et al., 2010). Marine waters are a large contributor of atmospheric N2O, accounting for 10–33% of the global source (Jiang et al., 2007). However the oceanic emissions are not uniformly distributed, and estuaries and coastal areas account for 35–60% of the total marine sources of N2O (Bange et al., 1998). In fact, the highest N2O saturations are found in estuarine systems, whereas shelf waters, which are not influenced by the freshwater plumes, are close to equilibrium with the atmosphere (Murray et al., 2015). This implies that N2O is mainly formed in estuarine systems (Bange, 2006). However it is important to validate these global estimates with local and regional measurements because of the considerable spatial heterogeneity that characterizes coastal areas, and because of the seasonal and inter-annual variability of air-water N2O exchange (Bange et al., 2009). Furthermore, areas influenced by urban effluents and primary sector activities such as agriculture and fish farming may make a relatively greater estuarine and coastal water contribution compared to pristine systems (Daniel et al., 2013; Rajkumar et al., 2008). The high variability of N2O fluxes from different coastal regions is highlighted in a recent review by Murray et al. (2015). Some examples are the Colde Estuary in England with a mean flux reaching 685 μmol m−2 day−1 (Robinson et al., 1998), the Seine Estuary in France with mean flux of 40 μmol m−2 day−1 (Garnier et al., 2006), as well as Asian estuaries like the Chanjiang Estuary in China with a mean flux of 87 μmol m−2 day−1 (Zhang et al., 2010). This paper presents the spatio-temporal variability of dissolved N2O in waters of three different coastal systems located in Cadiz Bay and the associated atmospheric fluxes. The main sources of N2O into the systems are also discussed and the total emission from the study area has been established. 2. Material and methods 2.1. Study site The three coastal systems studied are located to the east and south of Cadiz Bay (36° 29′ N, 6° 13′ W, Fig. 1). This temperate region (mean annual temperature of 18 °C) covers an area of 152 km2 and sustains a population of 42,500 inhabitants. Urban effluents and primary sector activities, particularly agriculture and fish farming, are sources of organic matter and nutrients to the systems. The Bay is characterized by semidiurnal meso-tides (average tidal range 0.98–3.20 m); therefore the water flow direction along the full length of the systems changes with incoming and outgoing tides. The first system, the estuary of the Guadalete River (G-Rv), extends to a length of 18 km and has a maximum depth of 6 m in the river mouth. The salinity gradient covers a range from about 35 in the River mouth to salinities close to 0 in the freshwater endmember. Eight sampling stations were established along the system; the first is located in

the mouth of the river and the last inland close to the effluent discharge point of the wastewater treatment plant serving Jerez de la Frontera (a city of 212,226 inhabitants, in 2014). Most of the terrain adjacent to the river is used for irrigated crops. In addition there are direct discharges of organic matter (OM) and nutrients from farms and individual households. The second system, the Rio San Pedro Creek (R-Cr), is located within a salt marsh area. It is characterized by a relatively constant depth of 3 m. The freshwater inflow into the system is not significant (except during seasonal rains) and the tidal current inundates the creek, therefore it is considered a marine system. Nine sampling stations were positioned along its 12 km length; the first is situated at its point of connection with the outer Bay. The R-Cr receives inputs of organic matter and nutrients from the wastewater discharges of a fish farm located between the last two stations furthest from the outlet. The upper part of the R-Cr receives the drainage of a large but very shallow salt marsh area. The third system is the Sancti Petri Channel (S-Ch) with eleven sampling stations; the first is located at the Atlantic outlet of the channel. This channel links the inner Bay with the Atlantic Ocean running some 18 km through the centre of a significant zone of salt marshes (6000 ha) flooded with seawater. The channel is deeper at both ends, north and south, reaching a maximum depth of 7 m, and very shallow in the central part (average depth of 2 m), where the Iro River mouth discharges freshwater between the sixth and seventh stations. This freshwater input does not lead to a salinity gradient because the tidal current inundates the channel from both ends and thus S-Ch is considered a marine system. The Iro River receives organic matter and nutrients from the sewage discharges of Chiclana de la Frontera (a town of 82,298 inhabitants, in 2014).

2.2. Sample collection Surface waters were sampled following three different sampling strategies with a view to studying the spatio-temporal variability of the measurements: i) two-year sampling at fixed stations; ii) longitudinal samplings conducted seasonally; and iii) daily samplings covering two tidal cycles. The positions of all the sampling points are given in Fig. 1. One fixed station located in each system was sampled twice per month from April 2013 to April 2015. These samplings were taken at a medium tidal coefficient (0.65 ± 0.04) and 3 h after the flood in order to minimise the tidal influence. Secondly, four longitudinal samplings were done seasonally, upstream of the systems, during 2013 (February, May, July and November); these were performed during the ebb tide with high tidal coefficients (between 0.82 and 1.06) with the object of capturing the maximum influence of the OM inputs and the in situ production of N2O. The third temporal strategy was daily samplings carried out in each of the three systems in July 2015, where surface waters were collected at intervals of 1 h. The locations were those of the three fixed stations. The object of the daily samplings was to investigate the influence of tides on the parameters studied. Current velocity was recorded with a current meter (Alec, EM) with an accuracy of ±1 cm s−1. Samples for dissolved nitrous oxide determinations were collected in 250 mL airtight glass bottles with frosted stopper, preserved with saturated mercuric chloride to inhibit microbial activity, and sealed with Apiezon® grease to prevent gas exchange. Samples for dissolved oxygen (DO) determination were collected in 250 mL airtight glass bottles and fixed. Frosted stopper glass jars of 250 mL were filled to measure nutrients (nitrite, nitrate, and ammonium). 5 L of surface water were collected in a plastic jug for particulate organic nitrogen (PON) measurements. All samples were stored in the dark until they were analyzed in the laboratory. A mercury thermometer recorded in situ temperature with an accuracy of ±0.1 °C, and salinity was quantified in the laboratory using an induction salinometer (Rousemount® Analytical) with an accuracy of ±0.001 units.

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

143

Fig. 1. Map of the Cadiz Bay (SW Spain) showing the sampling stations in the three systems studied: Guadalete River (G-Rv), Rio San Pedro Creek (R-Cr), and Sancti Petri Channel (S-Ch). Triangles show the location of the fixed stations and the daily samplings. Stars indicate points of wastewater discharge from anthropogenic activities.

2.3. Analytical methods Dissolved nitrous oxide was measured with a gas chromatograph (Bruker GC-450) using the headspace technique. The equilibration system is connected to the gas chromatograph to reduce sample manipulation, following Upstill-Goddard et al. (1996). A volume of about 70 mL of sample was equilibrated with a volume of about 180 mL of a gas standard with concentrations close to the atmospheric value. After equilibration at 25 °C, a sample of the gas space was injected into the gas chromatograph. A mixture of Ar/CH4 (95%/5%) is used as carrier gas (10 mL min−1), and gases were separated by a 0.5 m × 1/8-in stainless steel Hayesep N column (80/100) and 2.0 m × 1/8-in stainless steel Hayesep D column. Nitrous oxide was detected with an electron capture detector (ECD), operating at 350 °C. The detector was calibrated daily using three standard gases with concentrations of 0.3, 0.5, and 2.0 ppmv (certified by Abello Linde). The precision of the method based on replicate analysis (n = 64), including the equilibration step,

expressed as the coefficient of variation, was 4.5%. Seawater gas solubility was calculated using the function given by Weiss and Price (1980). The gas saturation, expressed in %, was calculated as the ratio of the concentration of the dissolved gas and the expected concentration in equilibrium with the atmosphere. The Winkler method has been applied for DO analysis (Grasshoff and Ehrhardt, 1983) using potentiometric endpoint titration (Metrohm, 702). Apparent oxygen utilization (AOU) was calculated using the solubility expression proposed by Weiss (1974). AOU is defined as the difference between measured O2 and the theoretical value of O2 in equilibrium with the atmosphere (AOU = [O2]measured − [O2]equilibrium). Similarly, the excess N2O (ΔN2O) is the difference between measured N2O and the N2O equilibrium value (ΔN2O = [N2O]measured − [N2O]equilibrium). Samples collected for determinations of nutrients and particulate organic nitrogen (PON) were filtered through precombusted glass fibre filters (Whatman, GF-F 0.7 μm). Nutrients were analyzed in a segmented flow autoanalyzer (Skalar, San Plus) based on classic

144

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

Table 1 Variation range of temperature, salinity, dissolved oxygen (DO), nutrients, and dissolved N2O, and mean values (±SD) of N2O saturation percentages (N2O Sat %) for all spatiotemporal samplings carried out in the three coastal systems: Guadalete River (G-Rv), Rio San Pedro Creek (R-Cr), and Sancti Petri Channel (S-Ch). T (°C)

S

DO (μM)

NO− 3 (μM)

NO− 2 (μM)

NH+ 4 (μM)

N2O (nM)

N2O Sat %

Fixed station G-Rv R-Cr S-Ch

11.8–25.9 10.7–26.5 10.5–26.0

22.4–37.0 25.1–39.2 23.9–42.1

108.1–263.1 86.8–173.2 74.2–273.0

2.0–180.8 0.3–26.0 2.8–38.6

0.51–10.35 0.05–1.60 0.38–2.50

6.4–107.0 1.6–25.1 3.4–25.8

6.1–34.7 4.1–17.8 4.1–19.8

188.0 ± 138.8 138.0 ± 56.3 158.9 ± 58.6

Longitudinal G-Rv R-Cr S-Ch

12.2–26.8 11.0–27.0 11.2–27.3

0.1–36.3 26.3–44.9 19.8–44.8

179.6–286.3 130.6–226.6 85.0–318.7

1.7–285.5 2.1–20.1 0.1–128.9

0.39–18.84 0.40–3.70 0.00–5.72

2.1–249.6 3.6–26.8 1.2–50.8

(43.0–292.0)a 39.1–76.2 35.2–126.8

1770.5 ± 864.0 793.8 ± 112.0 891.1 ± 323.3

Daily G-Rv R-Cr S-Ch

26.1–28.9 26.4–28.8 26.7–29.3

24.8–37.6 37.6–41.0 39.2–41.1

111.9–249.8 55.6–230.2 143.3–342.4

1.1–20.3 0.2–8.1 3.0–8.7

0.23–4.75 0.00–1.34 0.76–1.83

3.3–36.2 1.6–15.3 8.9–31.5

7.7–27.1 4.7–9.0 5.9–11.1

257.3 ± 96.5 109.2 ± 21.1 145.6 ± 25.5

a

Data from Burgos et al. (2015).

spectrophotometric methods (Grasshoff and Ehrhardt, 1983). PON was analyzed with a CHN (carbon-hydrogen-nitrogen) analyzer (LECO, 932) (Grasshoff and Ehrhardt, 1983). The accuracies of the determinations obtained are the following: ± 0.5 μmol kg−1 for dissolved oxygen; ± 0.02 μM for nitrite; ± 0.10 μM for nitrate; and ±0.05 μM for ammonium. The accuracy (given as a coefficient of variation) of the DON measurements was 2–3%. 2.4. Air-water flux calculation The water-air flux of N2O was determined following the equation proposed by Liss and Merlivat (1986), whereby the flux is estimated using the gas transfer velocity coefficient, k, and the concentration gradient between the surface water and the atmosphere. FN2 O ¼ k ðCw −αCa Þ

ð1Þ

where FN2O (μmol m−2 day−1) is the flux of nitrous oxide, k (cm h−1) is the gas transfer velocity, Cw is the concentration of N2O in the surface water (mol L− 1), α is the gas solubility in seawater given by Weiss and Price (1980), and Ca is the atmospheric concentration. A positive flux indicates a transfer of N2O from the systems to the atmosphere. The relationship between k and wind speed has been calculated using the relationship formulated by Raymond and Cole (2001) for estuarine environments. Daily mean wind speed and accumulated precipitation data was obtained from the meteorological station located close to the study site provided by the Spanish IFAPA Institute (Instituto de Investigación y Formación Agraria y Pesquera). N2O concentration in the atmosphere was assumed to correspond to the annual average concentration at the Mauna Loa Observatory for the years 2013, 2014, and 2015, respectively (NOAA Research, available at http://www.cmdl. gov/). The differences between these values and the local atmospheric concentrations are assumed to be small compared with the high dissolved concentrations in the surface waters of our study site.

Fig. 2. Distribution of dissolved N2O, apparent oxygen utilization (AOU), dissolved inorganic nitrogen (DIN) and particulate organic nitrogen (PON) monitored at the fixed station on the Guadalete River (G-Rv, black circles), Rio San Pedro Creek (R-Cr, grey triangles), and Sancti Petri Channel (S-Ch, white diamonds). The solid horizontal line represents the average value and dotted lines the standard deviation.

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

145

and spatial scale covered (Table 1). Results from the three sampling strategies (fixed stations, the longitudinal samplings carried out seasonally, and the daily cycles) are detailed in the following.

3.1. Fixed stations

Fig. 3. Monthly accumulated precipitation from January 2013 to April 2015 in the study region. Data has been provided by the Spanish IFAPA Institute (Instituto de Investigación y Formación Agraria y Pesquera, available at https://www.juntadeandalucia.es).

2.5. Statistical analysis Statistical analyses were performed using IBM SPSS Statistics software (Version 21.0. Armonk, NY). The one-way ANOVA test was used for analyzing significant differences of temporal and spatial variations of the dataset. Correlations between variables were established using Pearson's correlation test when the one-sample Kolmogorov-Smirnov test (K-S test) was not significant, and Spearman's correlation test when the K-S test was significant. The threshold value for statistical significance was taken to be p b 0.05.

3. Results Surface water properties (temperature, salinity, AOU, N2O and nutrients) showed different ranges of variation depending on the temporal

During two years of sampling, the temperature of the surface water varied between 10.5–11.8 °C in winter and 25.9–26.5 °C in summer (Table 1). Salinity ranged from 22.4 to 37.0 in the Guadalete River; however, the fixed station only represented the marine end-member of this estuary. The highest salinities were measured in the Rio San Pedro Creek and Sancti Petri Channel during summer, reaching values of 39.2 and 42.1, respectively, which may be the result of intense evaporation processes. DO varied between 74.2 and 342.4 μM in the whole study area, with a mean value (±SD) of 196.6 ± 45.8 μM, which indicates a well oxygenated water column. Negative AOU values were found in the three systems in a few samplings during spring and summer (Fig. 2), although most of the results showed positive AOU concentrations with average values of 44.1, 16.1, and 31.6 μM for G-Rv, R-Cr, and S-Ch, respectively. In general, N2O increased during the spring months. PON ranged from 0.05 to 0.41 mg L−1 with a mean value of 0.17 mg L−1 showing no significant differences among systems (p N 0.05, one-way ANOVA test). The greatest concentrations of N2O and PON were found during the year 2013 in the three environments, although no significant correlation has been established between these two variables. All nutrient concentrations (Table 1) and dissolved inorganic nitrogen (DIN) values (Fig. 2) were significantly higher in G-Rv than in the other two systems − (p b 0.05, one-way ANOVA test). NO− 3 and NO2 were positively correlated with the concentration of N2O (p b 0.05, Spearman's test). In addition, the correlation between DIN and N2O has been established in the two systems, G-Rv and R-Cr, whereas it was not observed in S-Ch. In average, the N2O saturation percentages were above 100% indicating that the water column was in general N2O supersaturated. None of the

Fig. 4. Longitudinal distribution of dissolved N2O and apparent oxygen utilization (AOU) in the three systems studied (Guadalete River G-Rv, Rio San Pedro Creek R-Cr, and Sancti Petri Channel S-Ch) during the year 2013.

146

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

Fig. 5. Longitudinal distribution of dissolved inorganic nitrogen (DIN) and particulate organic nitrogen (PON) in the three systems studied (Guadalete River G-Rv, Rio San Pedro Creek R-Cr, and Sancti Petri Channel S-Ch) during the year 2013.

described variables measured at the fixed stations showed significant differences among seasons (p N 0.05 one-way ANOVA test). The accumulated monthly precipitation varied seasonally during the timescale covered (Fig. 3). The highest rainfall took place in spring 2013 and fall 2014, reaching values above 200 mm. In winter the accumulated precipitation ranged between 50 and 80 mm. In contrast, precipitations were negligible during the summer months. 3.2. Seasonal variation along the longitudinal length of the systems The range of temperature measured seasonally in the longitudinal samplings was similar to that found at the fixed stations (Table 1).

However salinities changed over a wider range because of the longitudinal coverage of the systems. In G-Rv salinity increased from 0.1 in the freshwater end-member of the estuary to 36.3 in the river mouth. In S-Ch salinity decreased by 10 salinity units from the two outermost ends of the channel to the central part due to the influence of the Iro River discharge. During summer there was an inversion of this behaviour due to evaporation processes in the shallower central zone that result in an increase of salinity. Nevertheless, salinity fluctuation in R-Cr (26.3–44.9) and S-Ch (19.8–44.8) was more related to seasonal changes of precipitation/evaporation processes than to longitudinal variations. The AOU concentration was negative (with a minimum of −73.5 μM) in the outermost part of the systems, i.e. at the point of connection of

Table 2 Variation range of temperature, salinity, nutrients, and dissolved N2O and mean values (±SD) of N2O saturation percentages (N2O Sat %) for the longitudinal samplings made in the Guadalete River (G-Rv), Rio San Pedro Creek (R-Cr) and Sancti Petri Channel (S-Ch) during 2013. T (°C)

S

NO− 3 (μM)

NO− 2 (μM)

NH+ 4 (μM)

N2O (nM)

N2O Sat %

Winter G-Rv R-Cr S-Ch

12.2–14.3 11.0–12.7 11.2–13.4

1.1–34.2 26.3–35.1 19.8–34.7

17.0–285.5 8.2–20.7 5.9–128.9

0.49–17.59 0.58–2.85 0.48–5.72

3.2–202.3 3.6–16.4 2.3–22.0

(60.8–256.3)a 63.6–74.3 61.7–111.9

1260.8 ± 656.6 690.6 ± 26.0 791.0 ± 118.9

Spring G-Rv R-Cr S-Ch

20.7–22.1 20.4–21.8 18.8–22.7

0.9–30.2 30.7–34.6 33.2–36.3

78.3–250.1 6.7–10.1 0.1–25.1

4.83–10.61 0.72–2.12 0.14–4.85

34.4–126.8 4.4–15.1 1.2–50.8

(89.2–245.2)a 63.8–76.2 59.4–126.8

2309.3 ± 541.2 918.9 ± 64.1 1180.1 ± 385.7

Summer G-Rv R-Cr S-Ch

25.2–26.8 25.7–27.0 24.6–27.3

0.8–36.3 37.4–44.9 38.0–44.8

1.7–111.0 2.1–16.6 0.5–6.8

0.39–18.84 0.40–3.70 0.13–2.63

2.1–83.8 3.6–26.8 2.5–19.5

(43.0–292.0)a 39.1–51.4 35.2–61.5

1737.0 ± 1075.9 771.8 ± 76.7 693.2 ± 132.1

Fall G-Rv R-Cr S-Ch

13.8–14.9 12.4–14.3 13.3–15.4

0.1–30.1 36.7–39.2 37.0–38.5

17.9–99.1 3.1–8.6 0.9–16.2

1.75–4.30 0.49–1.02 0.10–2.74

30.8–246.6 5.4–8.4 1.7–43.9

– – –

– – –

a

Data from Burgos et al. (2015).

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

the systems with Cadiz Bay and also with the Atlantic Ocean in the case of S-Ch (Fig. 4). Upstream, AOU becomes positive, reaching values above 150 μM in the Guadalete River, and close to 50 μM in the Rio San Pedro Creek and Sancti Petri Channel. In these latter two systems, negative values of AOU were found in the innermost part during fall. In general, the highest concentrations of N2O were found in winter and spring. Unfortunately, N2O data from the fall samplings is not available. Dissolved N2O was significantly different (p b 0.05, one-way ANOVA test) between the outer and the inner part of the G-Rv and SCh systems. N2O increased further upstream in the G-Rv estuary, in parallel with the decrease of salinity (for more details, see Burgos et al., 2015). In S-Ch, N2O also increased in the proximity to the Iro River mouth. In contrast, the distribution of N2O did not show longitudinal + − variation in the R-Cr. All the nutrients (NO− 3 , N2O , and NH4 ), PON and DIN followed the same behaviour described for N2O within the systems. Nutrient concentrations were significantly higher in G-Rv than in R-Cr and S-Ch (p b 0.05, one-way ANOVA test). The concentration of N2O was positively correlated with the nutrients; the correlation of N2O with NO− 3 was strong in all three systems (p b 0.01, Spearman's test). In addition, N2O was strongly correlated with NH+ 4 (p b 0.01) in R-Cr and S-Ch. The concentration of PON was high in G-Rv during winter, reaching values up to 1.50 mg L−1 in the inner part of the estuary (Fig. 5). In S-Ch, the highest value of PON (0.50 mg L−1) was observed during summer. In contrast, the concentration of PON was low in R-Cr, and showed small longitudinal variation (mean ± SD was 0.11 ± 0.05). DIN varied

147

between 4.2 and 504.5 μM in G-Rv, between 6.1 and 46.4 μM in R-Cr, and between 4.2 and 150.3 μM in S-Ch, and showed a very close correlation with the concentration of N2O in the three systems (p b 0.01, Spearman's correlation test). PON and DIN were significantly correlated in G-Rv (p b 0.05, Spearman's correlation test). The maximum values of N2O saturation percentage were found in G-Rv (Table 2), in particular in the freshwater endmember, where concentration reached maximums of above 2500%. In the two marine systems, R-Cr and S-Ch, the N2O oversaturation was smoother with mean values of 793.8 and 891.1, respectively. Contrary to the result presented at the fixed stations, all the variables measured during the longitudinal samplings showed significant seasonal differences (p b 0.05 one-way ANOVA test). 3.3. Tidal variation All variables measured during the three daily samplings (summarized in Table 1 and Fig. 6) showed good correlation with the water height (p b 0.05, Pearson's test), with the exception of AOU in the Sancti Petri Channel. All variables were significantly different between low and high tide (p b 0.05, one-way ANOVA test). The temperature of the surface water was very high; it ranged from 26.1 to 29.3 °C, showing values above of those recorded in the previously described longitudinal samplings and at the fixed stations. Consequently, salinity was also high, reaching a maximum of 37.6 in G-Rv, and values above 40 in R-Cr and S-Ch. AOU varied in a wide range from negative to positive concentrations in the Rio San Pedro Creek (− 53.4–113.6 μM) and in the

Fig. 6. Tidal variation of dissolved N2O, apparent oxygen utilization (AOU), and dissolved inorganic nitrogen (DIN) in the three systems studied (Guadalete River G-Rv, Rio San Pedro Creek R-Cr, and Sancti Petri Channel S-Ch) during summer 2015.

148

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

Guadalete River (− 46.7–68.1 μM). In the Sancti Petri Channel AOU values were mainly positive (mean ± SD was 21.1 ± 22.0 μM). The highest concentration of N2O was measured in G-Rv, reaching values above 20 nM during low tide. Likewise, N2O showed the highest values during low tide in R-Cr and S-Ch, although concentrations were lower and close to 10 nM. The ranges of nutrient concentration were smaller in the daily sampling than in the longer time-scale samplings described above. Parallel to the daily distribution of N2O, the highest nutrient concentrations were associated with low tide. In particular, NO− 2 was very low during the ebb in R-Cr, coinciding with very low concentrations of + NO− 3 and NH4 (0.2 and 1.6 μM, respectively). DIN varied between 4.8 and 60.4 μM in G-Rv; between 1.9 and 23.6 μM in R-Cr; and between 13.6 and 37.8 in S-Ch. Similarly, this variable reached its maximum values during the low tide. 3.4. Air-water exchange Surface waters of the three ecosystems studied showed an average oversaturation of N2O with respect to the atmospheric equilibrium across the whole spatio-temporal range covered (Table 3). Nevertheless the lower limit of the range indicates that a few of the sampling points present b100% saturation. Wind speed was moderate in the longitudinal and daily samplings (1.3–2.2 m s−1) and the highest velocities appeared at the fixed stations (1.1–5.9 m s− 1). FN2O was, on average, positive, although a few measurements coinciding with spring and summer months showed weakly negative fluxes at the fixed stations and in the daily samplings of R-Cr and S-Ch. Significant differences of FN2O appeared among the systems (p b 0.01). In addition, fluxes increased with the proximity to the discharges of OM and nutrients in the Guadalete River and Sancti Petri Channel, reaching maximum values of 313.2 and 97.8 nmol m−2 day−1, respectively. 4. Discussion 4.1. Spatiotemporal variability of N2O distribution Concentrations of N2O in surface water obtained from the different temporal and spatial sampling strategies undertaken in the three coastal systems ranged between 1.1 and 292.0 nM, and showed wide variability. These values are within the wide range previously observed in a number of other coastal systems (see, for instance, the lists compiled in Bange, 2006 and Murray et al., 2015), that also show considerable variability. The most important factors controlling N2O distribution in coastal systems seem to be the temperature and the availability of DIN and oxygen (Murray et al., 2015). Table 3 Wind speed (WS, averages for the day of sampling), N2O saturation percent and sea-air N2O flux for all the spatiotemporal measurements in the three systems studied: Guadalete River (G-Rv), Rio San Pedro Creek (R-Cr), and Sancti Petri Channel (S-Ch). WS (m s−1)

Saturation percent

Flux (μmol m−2 day−1)

Range

Mean ± SD

Range

Mean ± SD

Range

Fixed station G-Rv 1.1–5.9 R-Cr 1.1–5.9 S-Ch 1.1–5.9

194 ± 137 142 ± 52 163 ± 56

29–556 71–233 59–240

11.3 ± 7.1 2.4 ± 3.4 3.6 ± 4.0

0.5–28.8 −5.6–7.1 −6.2–13.4

Longitudinal 1.4–2.1 G-Rva R-Cr 1.4–2.5 S-Ch 1.3–2.2

1300 ± 864 793 ± 112 891 ± 323

663–2714 619–1025 579–1781

124.4 ± 69.8 46.4 ± 5.2 49.2 ± 19.1

35.0–313.2 33.0–55.2 30.1–97.8

Daily G-Rv R-Cr S-Ch

257 ± 69 109 ± 21 146 ± 25

123–419 80–157 90–195

9.6 ± 6.2 0.4 ± 1.1 2.9 ± 1.5

1.3–20.2 −1.1–2.9 −0.2–5.9

a

1.9 1.7 2.2

Data from Burgos et al. (2015).

Dissolved N2O results from the fixed stations tended to increase in spring coinciding with the availability of more DIN (r2 = 0.52). This finding may be related to the seasonal rainfall (Fig. 3) since, in general, dissolved N2O and DIN increased after the months with high accumulated precipitation, and they decreased during the dry months of summer. The seasonal trend was masked in the fixed station dataset possibly due to dilution by tidal action. In contrast all the variables measured in the longitudinal sampling showed significant differences among seasons, since these samplings were done during the ebb tide and, therefore, the concentration gradients were stronger. The increase of N2O in spring was also observed in the longitudinal samplings (Fig. 5). Rainfall increases the OM and nutrients load from both riverine discharge and anthropogenic sources, particularly agriculture, sewage treatment plants, and fish farming. Worldwide, many coastal marine ecosystems are receiving increased nitrogen from such anthropogenic activities, which boosts marine production of N2O (Barnes and Upstill-Goddard, 2011; Galloway et al., 2003; Musenze et al., 2014; Naqvi et al., 2000; Zhang et al., 2006). In contrast, Ferron et al. (2007) found that seasonality of N2O in the Rio San Pedro Creek during 2004 was closely linked to the temperature cycle, showing the highest N2O concentrations during summer. The inversion of the annual trend observed in our N2O results in R-Cr is probably associated with the significant reduction of fish farming activity in the last 10 years; this commercial activity currently releases less wastewater than previously reported (Tovar et al., 2000). The highest concentrations of N2O, together with the availability of more nutrients, were found in the inner estuary of the G-Rv. The inner estuarine concentrations of N2O are typically up to an order of magnitude higher than those found in outer estuaries (Barnes and UpstillGoddard, 2011). Furthermore, dissolved N2O varied longitudinally in G-Rv and S-Ch but not in R-Cr (Fig. 5). In the estuary of the G-Rv, the concentration of N2O increased upstream, linked to the salinity gradient, and there were statistically significant differences between values at the river mouth and those at the freshwater end-member (p b 0.05). Dissolved N2O reached values above 200 nM in the inner part that can be explained by the river discharge and proximity to where effluent is released from the wastewater treatment plant. The longitudinal variation in S-Ch was also related to the Iro River discharges in the central part of the channel, which, in turn, receives wastewater from the sewage treatment plant of Chiclana de la Frontera. In contrast, the distribution of N2O in R-Cr seems to be a result of both low N2O input and production in this system and the action of tides that spread the dissolved biogas along the creek and thus no evident gradient was observed. Results from the two tidal cycles indicate a strong variability of the concentration of N2O with tides (Fig. 6). Dissolved N2O increased when tidal elevation decreased, with maximum values during the ebb. The same N2O variation with tides has been found in other studies (Barnes and Upstill-Goddard, 2011; Ferron et al., 2007; Gonçalves et al., 2010). This pattern suggests that the inner waters of the three coastal systems contain a higher concentration of N2O than the waters of the Cadiz Bay and the Atlantic Ocean. Several authors have pointed out that benthic mineralization can be a major source of N2O to the water column in different coastal systems (Dong et al., 2002; Meyer et al., 2008; Zhang et al., 2010). For instance, Ferron et al. (2009) reported benthic fluxes of N2O in R-Cr that ranged seasonally between 4.3 and 49 μmol m−2 day−1. During the ebb, the sediments as a source of N2O may be more significant due to the water column having a smaller volume. In addition, sediment pore water may release more dissolved N2O in low tide because of hydrostatic pressure drop toward low water. Another possible source of N2O is the lateral transport from adjacent salt marshes through tidal exchange. It is notable that the variation range of dissolved N2O due to tides was close to that found at the fixed stations (Table 1). We can conclude that the most significant variability has been found in the longitudinal samplings associated with the seasonal rainfall. This suggests that the

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

main source of N2O into these systems was probably inputs from anthropogenic activity effluents and adjacent salt marshes, and the N2O production processes taking place within the systems.

4.2. Reactivity in the water column There seems to be a diversity of sources contributing nitrogen into the systems studied. The significant direct correlation found between PON and DIN in G-Rv (p b 0.05, Spearman's test) suggests that mineralization of the organic matter may be a source of DIN species within this system. The contribution of NH+ 4 to DIN varied between 30 and 72% while the contribution of NO− 3 was between 23 and 66%. Surface − water concentrations of NH+ 4 and NO3 in G-Rv were high (on average, 45.0 and 57.1 μM, respectively) indicating anthropogenic influence on the N budget of the estuary. For comparison, concentrations of NH+ 4 and NO− 3 in rivers considered pristine are approximately 2 μM for − NH+ 4 and 8 μM for NO3 (Meybeck and Ragu, 1995). Concentrations of both nutrients were on average lower in R-Cr and S-Ch (Table 1), probably due to lower inputs from anthropogenic activities and the process by which the waters from the systems mix with the marine source. The mixing process may also explain the absence of correlation between PON and DIN in these two systems.

149

Current knowledge concerning N2O formation in seawater is that there are only two dominant microbial processes: nitrification and denitrification (Bange et al., 2010). − During nitrification, nitrifying bacteria oxidize NH+ 4 to NO3 in the presence of O2, and produce N2O as a byproduct (Ward, 2008). In low oxygen/anoxic environments, denitrifying bacteria reduce NO− 3 to N2, and N2O is an intermediate product (Capone et al., 2008). Our measurements found the water column always oxygenated, which indicates only limited possibilities for water column denitrification. Nevertheless, denitrification has been identified as the main process responsible for N2O production in sediments (Meyer et al., 2008), which may, in turn, be a source of N2O to the water column. Excess N2O (ΔN2O) is defined as the difference between measured N2O and the theoretical N2O equilibrium value. AOU is a measure of the amount of O2 consumed during organic matter mineralization in the water. Because nitrification is part of the OM mineralization sequence, plots of ΔN2O versus AOU have been used to identify the prevailing processes of generation and consumption of N2O in the water column (e.g. Gonçalves et al., 2012; Nevison et al., 2003; Yevenes et al., 2016). Therefore positive linear ΔN2O – AOU relationships suggest that nitrification is the main process generating N2O in seawater (Bange and Andreae, 1999). The ΔN2O–AOU correlation has been found in the longitudinal samplings (r2 = 0.53) (Fig. 7). The slope obtained by linear

Fig. 7. Excess N2O (ΔN2O) versus apparent oxygen utilization (AOU) and nitrate concentration in the Guadalete River (G-Rv, black circles), Rio San Pedro Creek (R-Cr, grey triangles), and Sancti Petri Channel (S-Ch, white diamonds) over the entire spatio-temporal dataset. The linear correlation coefficients are shown in each plot and the linear fit when r2 N 0.5.

150

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151

correlation analysis for ΔN2O versus AOU was 0.75 nmol μmol−1. This ratio reflects the N2O yield per mol of O2 consumed. Our ratio is an order of magnitude higher than those found in estuarine and oceanic sites (Nevison et al., 2003; Yevenes et al., 2016), indicating that dissolved N2O in the systems studied was affected by other processes besides the water column production by nitrification (such as benthic fluxes, lateral inputs and anthropogenic discharges). In some studies a caveat has been issued against a straightforward interpretation of the linear ΔN2O–AOU relationship as an indicator for N2O formation via nitrification, because this linear correlation may not necessarily result from nitrification alone (Nevison et al., 2003; Yamagishi et al., 2005). Additionally, plots of ΔN2O versus NO− 3 have also been used to identify N2O formation through nitrification since in the process NO− 3 is produced along with N2O (Walter et al., 2006; Yevenes et al., 2016). The correlation has been established in all the spatio-temporal samplings (Fig. 7). The R square was 0.44 at the fixed stations, 0.75 in the longitudinal samplings and 0.83 in the daily samplings, which are yet more evidence of N2O formation in the water column by nitrification. The slope of the ΔN2O–NO− 3 linear correlation varied between 0.17 and 1.02 nmol μmol−1 for all the samplings. These values are close to that found by Yevenes et al. (2016) in the Reloncaví Estuary in the Chilean Patagonia (0.27–0.74 nmol μmol−1). 4.3. N2O emission to the atmosphere The study area was on average supersaturated with N2O during all the spatio-temporal samplings (Table 3), indicating that the three systems were a source of N2O to the atmosphere. The flux between the surface waters and the atmosphere depends mainly on the concentration gradient and the gas transfer velocity, k. For gases with low solubility, such as N2O, k mainly depends on the turbulence at the aqueous boundary layer. In lakes and in the ocean, wind is the major forcing factor generating turbulence, and k is often parameterized as a function of wind speed at 10 m height (e.g. Ho et al., 2006; Nightingale et al., 2000; Wanninkhof, 1992). In the case of shallow coastal sites, turbulence in the aqueous boundary layer is additionally generated by the friction of water flow at the bottom, and k depends on water current velocity, water depth, and bed roughness (Borges et al., 2004; O'Connor and Dobbins, 1958), and there is a lack of agreement on the relationship proposed in the literature. The parameterization of k from Raymond and Cole (2001) is derived specifically for estuaries but neglecting bottomgenerated turbulence. For example, Ho et al. (2014) found that, by neglecting bottom generated turbulence, k600 was underestimated by between 40 and 70% in the Shark River estuary. Nevertheless the parameterization proposed by Raymond and Cole (2001) has been widely used in calculating gas exchange in coastal systems (e.g. Biswas et al., 2004; Guérin et al., 2007; Musenze et al., 2014). Consistent with the saturation values, the estimated N2O fluxes in our study area were on average positive (Table 3). Moreover, they showed spatio-temporal variability. FN2O did not show correlation with the temperature, indicating that the temperature influence on gas solubility was masked by the inputs of N2O to the systems and its production within the water column. Therefore, the distribution of FN2O along the systems showed a trend very similar to that discussed for the concentration of nitrous oxide in surface waters. This fact also suggests that, at low wind speed (1.3–2.5 m s−1), fluxes are mainly driven by the concentration gradient between the water and the atmosphere. G-Rv was a strong source of N2O, reaching values above 300 μmol m−2 day−1 in the inner estuary; those values are comparable to the N2O emissions from other estuaries affected by anthropogenic activities (Daniel et al., 2013; Rajkumar et al., 2008; Tong et al., 2013). FN2O in R-Cr are consistent to the range previously found by Ferron et al. (2007) in the same Creek (ranging between 24 and 62 μmol m−2 day−1). Average fluxes in R-Cr and S-Ch were statistically similar (46 and 49 μmol m−2 day−1, respectively; p b 0.05, one-way ANOVA test) and in good agreement with values found worldwide in temperate salt marshes (Adams et al., 2012; Hirota et al., 2007;

Sun et al., 2014). The mean flux considering the three coastal systems investigated was 69 μmol m− 2 day−1, which corresponds to a total emission of 240 · 103 mol of N2O year−1. 5. Conclusions This paper discusses the distribution of dissolved N2O and its fluxes to the atmosphere in three shallow coastal systems on different time scales. Results highlight the substantial heterogeneity of N2O distribution in the surface waters. Seasonal variation seems to be linked to the precipitation regime, which in turns increases the OM and nutrient load from anthropogenic activities, particularly urban effluents, agriculture and fish farming, and lateral inputs from adjacent salt marshes. A major part of the variability of N2O was also associated with the tides. Very high concentrations found during the ebb indicate that the systems export dissolved N2O to the Cadiz Bay and the Atlantic Ocean. Sediments may be a source of N2O to the water column, which may also contribute to explaining the distribution of N2O. Longitudinal variation revealed an increase of N2O toward the inner part of the Guadalete River and Sancti Petri Channel associated with river discharges that bring wastewater from sewage treatment plants. In contrast, longitudinal variation was not observed in the Rio San Pedro Creek, which may be related to a lower level of N2O production, reduced inputs into this system, and higher water renovation as a result of the action of tides. Plots of AOU versus ΔN2O (N2O excess) and NO− 3 versus ΔN2O indicate that nitrification is an important process for in situ formation of N2O. The study area was in average a net source of N2O to the atmosphere with fluxes ranging for the whole spatiotemporal monitoring between 0.5 and 313.2 μmol m− 2 day−1 in GRv; − 7.2 and 55.2 μmol m− 2 day−1 in R-Cr; and − 6.2 and 97.8 μmol m−2 day−1 in S-Ch. Acknowledgements This work was supported by the Spanish CICYT (Spanish Program for Science and Technology) under contracts CTM2011-27891 and CTM2014-59244-C3-1-R. M. Burgos was financed by the Spanish Ministry of Education with a FPI fellowship (BES-2012-053882). References Adams, C.A., Andrews, J.E., Jickells, T., 2012. Nitrous oxide and methane fluxes vs. carbon, nitrogen and phosphorous burial in new intertidal and saltmarsh sediments. Sci. Total Environ. 434, 240–251. Bange, H.W., 2006. Nitrous oxide and methane in European coastal waters. Estuar. Coast. Shelf Sci. 70:361–374. http://dx.doi.org/10.1016/j.ecss.2006.05.042. Bange, H.W., Andreae, M.O., 1999. Nitrous oxide in the deep waters of the world's oceans. Glob. Biogeochem. Cycles 13, 1127–1135. Bange, H.W., Dahlke, S., Ramesh, R., Meyer-Reil, L.A., Rapsomanikis, S., Andreae, M.O., 1998. Seasonal study of methane and nitrous oxide in the coastal waters of the southern Baltic Sea. Estuar. Coast. Shelf Sci. 47, 807–817. Bange, H.W., Bell, T.G., Cornejo, M., Freing, A., Uher, G., Upstill-Goddard, R.C., Zhang, G., 2009. MEMENTO: a proposal to develop a database of marine nitrous oxide and methane measurements. Environ. Chem. 6, 195–197. Bange, H.W., Freing, A., Kock, A., Löscher, C.R., 2010. Marine pathways to nitrous oxide. Nitrous oxide. Clim. Chang. 36–62. Barnes, J., Upstill-Goddard, R.C., 2011. N2O seasonal distributions and air-sea exchange in UK estuaries: implications for the tropospheric N2O source from European coastal waters. J. Geophys. Res. Biogeosci. 116. http://dx.doi.org/10.1029/2009JG001156. Biswas, H., Mukhopadhyay, S.K., De, T.K., Sen, S., Jana, T.K., 2004. Biogenic controls on the air—water carbon dioxide exchange in the Sundarban mangrove environment, northeast coast of Bay of Bengal, India. Limnol. Oceanogr. 49, 95–101. Borges, A.V., Vanderborght, J.-P., Schiettecatte, L.-S., Gazeau, F., Ferrón-Smith, S., Delille, B., Frankignoulle, M., 2004. Variability of the gas transfer velocity of CO2 in a macrotidal estuary (the Scheldt). Estuaries 27, 593–603. Burgos, M., Sierra, A., Ortega, T., Forja, J.M., 2015. Anthropogenic effects on greenhouse gas (CH4 and N2O) emissions in the Guadalete River Estuary (SW Spain). Sci. Total Environ. 503–504:179–189. http://dx.doi.org/10.1016/j.scitotenv.2014.06.038. Capone, D.G., Bronk, D.A., Mulholland, M.R., Carpenter, E.J., 2008. Nitrogen in the Marine Environment. Academic Press. Chong, L.S., Prokopenko, M.G., Berelson, W.M., Townsend-Small, A., McManus, J., 2012. Nitrogen cycling within suboxic and anoxic sediments from the continental margin of Western North America. Mar. Chem. 128–129:13–25. http://dx.doi.org/10.1016/j.marchem.2011. 10.007.

M. Burgos et al. / Science of the Total Environment 607–608 (2017) 141–151 Dalal, R.C., Wang, W., Robertson, G.P., Parton, W.J., 2003. Nitrous oxide emission from Australian agricultural lands and mitigation options: a review. Soil Res. 41, 165–195. Daniel, I., DeGrandpre, M., Farías, L., 2013. Greenhouse gas emissions from the Tubul-Raqui estuary (central Chile 36°S). Estuar. Coast. Shelf Sci. 134, 31–44. Dong, L.F., Nedwell, D.B., Underwood, G.J.C., Thornton, D.C.O., Rusmana, I., 2002. Nitrous oxide formation in the Colne estuary, England: the central role of nitrite. Appl. Environ. Microbiol. 68, 1240–1249. Ferron, S., Ortega, T., Gómez-Parra, A., Forja, J.M., 2007. Seasonal study of dissolved CH4, CO2 and N2O in a shallow tidal system of the bay of Cádiz (SW Spain). J. Mar. Syst. 66:244–257. http://dx.doi.org/10.1016/j.jmarsys.2006.03.021. Ferron, S., Ortega, T., Forja, J.M., 2009. Benthic fluxes in a tidal salt marsh creek affected by fish farm activities: Río San Pedro (Bay of Cádiz, SW Spain). Mar. Chem. 113:50–62. http://dx. doi.org/10.1016/j.marchem.2008.12.002. Galloway, J.N., Aber, J.D., Erisman, J.W., Seitzinger, S.P., Howarth, R.W., Cowling, E.B., Cosby, B.J., 2003. The nitrogen cascade. Bioscience 53, 341–356. Garnier, J., Cébron, A., Tallec, G., Billen, G., Sebilo, M., Martinez, A., 2006. Nitrogen behaviour and nitrous oxide emission in the tidal Seine River estuary (France) as influenced by human activities in the upstream watershed. Biogeochemistry 77, 305–326. Gonçalves, C., Brogueira, M.J., Camões, M.F., 2010. Seasonal and tidal influence on the variability of nitrous oxide in the Tagus estuary, Portugal. Sci. Mar. 74:57–66. http://dx.doi.org/10. 3989/scimar.2010.74s1057. Gonçalves, C., Brogueira, M.J., Camões, M.F., 2012. Nitrous oxide content and fluxes in Setúbal Bay, Portugal, during upwelling events (Concentraciones y flujos de óxido de nitrógeno disuelto en la bahía de Setúbal, Portugal, durante los eventos de surgencia). Cienc. Mar. 38, 177–190. Grasshoff, K., Ehrhardt, M., 1983. Automated chemical analysis. Methods of Seawater Analysis. Verlag, pp. 263–289. Guérin, F., Abril, G., Serça, D., Delon, C., Richard, S., Delmas, R., Tremblay, A., Varfalvy, L., 2007. Gas transfer velocities of CO 2 and CH 4 in a tropical reservoir and its river downstream. J. Mar. Syst. 66, 161–172. Hirota, M., Senga, Y., Seike, Y., Nohara, S., Kunii, H., 2007. Fluxes of carbon dioxide, methane and nitrous oxide in two contrastive fringing zones of coastal lagoon, Lake Nakaumi, Japan. Chemosphere 68:597–603. http://dx.doi.org/10.1016/j.chemosphere.2007.01.002. Ho, D.T., Law, C.S., Smith, M.J., Schlosser, P., Harvey, M., Hill, P., 2006. Measurements of air-sea gas exchange at high wind speeds in the Southern Ocean: implications for global parameterizations. Geophys. Res. Lett. 33. Ho, D.T., Ferrón, S., Engel, V.C., Larsen, L.G., Barr, J.G., 2014. Air-water gas exchange and CO2 flux in a mangrove-dominated estuary. Geophys. Res. Lett. 41:108–113. http://dx.doi.org/10. 1002/2013GL058785. IPCC, 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change Stocker. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Jiang, X., Ku, W.L., Shia, R., Li, Q., Elkins, J.W., Prinn, R.G., Yung, Y.L., 2007. Seasonal cycle of N2O: analysis of data. Glob. Biogeochem. Cycles 21. Liss, P.S., Merlivat, L., 1986. Air-sea gas exchange rates: introduction and synthesis. The Role of Air-Sea Exchange in Geochemical Cycling. Springer, pp. 113–127. Meybeck, M., Ragu, A., 1995. River Discharges to the Oceans: An Assessment of Suspended Solids, Major Ions and Nutrients. UNEP. Meyer, R.L., Allen, D.E., Schmidt, S., 2008. Nitrification and denitrification as sources of sediment nitrous oxide production: a microsensor approach. Mar. Chem. 110:68–76. http://dx.doi. org/10.1016/j.marchem.2008.02.004. Middelburg, J.J., Klaver, G., Nieuwenhuize, J., Markusse, R.M., Vlug, T., Nat, F.J.W.A., 1995. Nitrous oxide emissions from estuarine intertidal sediments. Hydrobiologia 311, 43–55. Murray, R.H., Erler, D.V., Eyre, B.D., 2015. Nitrous oxide fluxes in estuarine environments: response to global change. Glob. Chang. Biol. http://dx.doi.org/10.1111/gcb.12923 (n/a-n/a.).

151

Musenze, R.S., Werner, U., Grinham, A., Udy, J., Yuan, Z., 2014. Methane and nitrous oxide emissions from a subtropical estuary (the Brisbane River estuary, Australia). Sci. Total Environ. 472, 719–729. Naqvi, S.W., Jayakumar, D.A., Narvekar, P.V., Naik, H., Sarma, V.V., D'Souza, W., Joseph, S., George, M.D., 2000. Increased marine production of N2O due to intensifying anoxia on the Indian continental shelf. Nature 408:346–349. http://dx.doi.org/10.1038/35042551. Nevison, C., Butler, J.H., Elkins, J.W., 2003. Global distribution of N2O and the ΔN2O-AOU yield in the subsurface ocean. Glob. Biogeochem. Cycles 17. Nightingale, P.D., Liss, P.S., Schlosser, P., 2000. Measurements of air-sea gas transfer during an open ocean algal bloom. Geophys. Res. Lett. 27:2117–2120. http://dx.doi.org/10.1029/ 2000GL011541. O'Connor, D.J., Dobbins, W.E., 1958. Mechanism of reaeration in natural streams. Trans. Am. Soc. Civ. Eng. 123, 641–666. Rajkumar, A., Barnes, J., Ramesh, R., Purvaja, R., Upstill-Goddard, R.C., 2008. Methane and nitrous oxide fluxes in the polluted Adyar River and estuary, SE India. Mar. Pollut. Bull. 56, 2043–2051. Raymond, P.A., Cole, J.J., 2001. Technical notes and comments: gas exchange in rivers and estuaries: choosing a gas transfer velocity. Estuaries 24, 312–317. Robinson, A.D., Nedwell, D.B., Harrison, R.M., Ogilvie, B.G., 1998. Hypernutrified estuaries as sources of N2O emission to the atmosphere: the estuary of the River Colne, Essex, UK. Mar. Ecol. Prog. Ser. 164:59–71. http://dx.doi.org/10.3354/meps164059. Sun, Z., Wang, L., Mou, X., Jiang, H., Sun, W., 2014. Spatial and temporal variations of nitrous oxide flux between coastal marsh and the atmosphere in the Yellow River estuary of China. Environ. Sci. Pollut. Res. 21:419–433. http://dx.doi.org/10.1007/s11356-013-1885-5. Tong, C., Huang, J.F., Hu, Z.Q., Jin, Y.F., 2013. Diurnal variations of carbon dioxide, methane, and nitrous oxide vertical fluxes in a subtropical estuarine marsh on neap and spring tide days. Estuar. Coasts 36, 633–642. Tovar, A., Moreno, C., Manuel-Vez, M.P., Garcı́a-Vargas, M., 2000. Environmental implications of intensive marine aquaculture in earthen ponds. Mar. Pollut. Bull. 40, 981–988. Upstill-Goddard, R.C., Rees, A.P., Owens, N.J.P., 1996. Simultaneous high-precision measurements of methane and nitrous oxide in water and seawater by single phase equilibration gas chromatography. Deep-Sea Res. I Oceanogr. Res. Pap. 43, 1669–1682. Walter, S., Bange, H.W., Breitenbach, U., Wallace, D.W.R., 2006. Nitrous oxide in the North Atlantic Ocean. Biogeosci. Discuss. 3, 993–1022. Wanninkhof, R., 1992. Relationship between wind speed and gas exchange over the ocean. J. Geophys. Res. Oceans 97, 7373–7382. Ward, B.B., 2008. Nitrification in marine systems. Nitrogen. Mar. Environ. 5, 199–261. Weiss, R.F., 1974. Carbon dioxide in water and seawater: the solubility of a non-ideal gas. Mar. Chem. 2, 203–215. Weiss, R.F., Price, B.A., 1980. Nitrous oxide solubility in water and seawater. Mar. Chem. 8: 347–359. http://dx.doi.org/10.1016/0304-4203(80)90024-9. Yamagishi, H., Yoshida, N., Toyoda, S., Popp, B.N., Westley, M.B., Watanabe, S., 2005. Contributions of denitrification and mixing on the distribution of nitrous oxide in the North Pacific. Geophys. Res. Lett. 32. Yevenes, M.A., Bello, E., Sanhueza-Guevara, S., Farías, L., 2016. Spatial distribution of nitrous oxide (N2O) in the Reloncaví Estuary–sound and adjacent sea (41°–43° S), Chilean Patagonia. Estuar. Coasts http://dx.doi.org/10.1007/s12237-016-0184-z. Zhang, G., Zhang, J., Xu, J., Zhang, F., 2006. Distributions, sources and atmospheric fluxes of nitrous oxide in Jiaozhou Bay. Estuar. Coast. Shelf Sci. 68:557–566. http://dx.doi.org/10. 1016/j.ecss.2006.03.007. Zhang, G.-L., Zhang, J., Liu, S.-M., Ren, J.-L., Zhao, Y.-C., 2010. Nitrous oxide in the Changjiang (Yangtze River) Estuary and its adjacent marine area: riverine input, sediment release and atmospheric fluxes. Biogeosciences 7, 3505–3516.